Impact of reconstruction parameter settings of knowledge-based iterative model technique on image quality of low-dose hepatic contrast-enhanced CT / 中国医学影像技术

ABSTRACT

Objective To evaluate the impact of different reconstruction parameter setting of knowledge-based iterative model reconstruction (IMR) technique on image quality of low-dose hepatic contrast-enhanced CT.Methods Forty patients underwent hepatic contrast-enhanced CT scanning were enrolled.Plain CT and triphasic contrast-enhanced CT scans,including hepatic arterial phase,portal-venous phase and delayed phase were performed.Low-dose scan was used in delayed phase,with tube voltage of 80 kV and tube current of 150 mAs.Images of delayed phase were reconstructed with both filtered back projection (FBP) and IMR techniques.Parameter setting applied in IMR reconstructions consisted of body routine and body soft tissue modes with three levels (Level 1-3),so 6 subgroups (R1,R2,R3,S1,S2,S3) were included.Subjective and objective evaluations of image quality were compared among those groups.Subjective evaluations included the scores of low contrast detectability (LCD),image distortion (ID) and diagnostic confidence (DC).Objective evaluations included image noise,signal to noise ratio (SNR) and contrast to noise ratio (CNR).Results There were statistical differences of scores in LCD,ID and DC among all the reconstruction parameter setting groups (all P＜0.01).The noise,SNR and CNR among different parameter setting groups had statistically significant differences (all P＜0.01).Except for subgroup S1 and subgroup R2,subgroup S2 and subgroup R3 (all P＞0.05),the other multiple comparisons showed significant differences (all P＜0.01).Conclusion IMR can improve image quality of low-dose hepatic contrast-enhanced CT.IMR reconstruction parameter setting of S1 and R2 are optimal for low-dose hepatic contrast-enhanced CT protocal.